Psychology, Crime & Law

Vol. 15, No. 8, October 2009, 743
758

The relationship between antisocial stereotypes and public CCTV

systems: exploring fear of crime in the modern surveillance society
Dave Williams* and Jobuda Ahmed

Department of Psychology, University of Hertfordshire, College Lane, Hatfield AL10 9AB, UK

(Received 19 October 2007; final version received 5 November 2008)

Situational crime deterrence measures like CCTV are not always associated with
reductions in fear of crime. This study explores this unexpected finding by
investigating the interaction between target type and the presence of a CCTV
camera, in order to test the effect this has on impressions of the target and
corresponding fear of the location the target was shown in. Participants (n 120)
were shown either a picture of a male ‘skinhead’, a ‘studious’ female, or no one
within an urban setting in which an obvious CCTV camera was either present or
absent. Participants then rated the scene using scales estimating crime frequency,
worry and target activity. Estimates of location safety fell for the male ‘skinhead’
target and activity impressions were more negative, but only when a CCTV
camera was also present. Ironically, in some circumstances, public crime
deterrence measures may prime pre-existing negative stereotypes about others
and so foster suspicion, undermine trust in others, and increase fear of crime.
Keywords: CCTV; surveillance; fear of crime; gender stereotypes; trust

Introduction
Open street public CCTV systems are an ever-present feature of the modern urban
landscape. It has been estimated that there are some 4.2 million cameras that are
either within, or visible from public space in the UK (McCahill & Norris, 2003;
Norris, McCahill, & Wood, 2004). CCTV systems are designed to serve a number of
goals (Gill & Spriggs, 2005; Williams, 2008), with reductions in fear of crime (FOC)
being a stated objective since the early days of their introduction (Honess &
Charman, 1992, p. 2). More recently, Gill and Spriggs (2005, p. 44) describe reducing
FOC as one of CCTV’s ‘main objectives’. However, the findings in relation to CCTV
and public reassurance are complex and mixed. For example, Ditton (2000) found
that respondents prospectively estimated a reduction in FOC (safety) after the
introduction of CCTV to Glasgow city centre. Despite this, when the system was
installed no significant difference emerged in FOC, a pattern that paralleled results
in two control locations. In fact, somewhat strikingly, Ditton also found that
respondent claims that they would ‘avoid’ the city centre (where cameras had been
installed) rose consistently from 50 to 59 and 65% in the years 1994, 1995 and 1996,
respectively. Yet, over the same period respondents showed the reverse trend in
control locations without cameras (43, 39 and 37%). These sorts of findings are not

*Corresponding author. Email: d.a.williams@herts.ac.uk

ISSN 1068-316X print/ISSN 1477-2744
# 2009 Taylor & Francis
DOI: 10.1080/10683160802612882
http://www.informaworld.com
744 D. Williams and J. Ahmed

confined to a single measure of FOC within the Ditton (2000) study, nor are they
confined to this study alone. More recently Gill and Spriggs (2005) sought to test
whether people’s worry about being a victim of crime, or their felt safety, altered after
the introduction of CCTV systems in 12 different locations. They found a reduction
in victimization worry in all locations after the installation of CCTV, but only in two
of these locations was the effect significantly different from control locations. A
further comparison between individuals within two areas was undertaken to test if
those who were more aware of the cameras showed less worry about crime. The
results showed the opposite effect, as: ‘those respondents who were aware of the
cameras actually worried more often about crime than those who were not’ (p. 48).
Gill and Spriggs (2005) claim this may be due to the way in which the assumed need
for CCTV in an area makes the location seem more problematic than was previously
thought. Some studies do attribute reductions in FOC to CCTV, but often this main
finding is restricted by further analysis. For example, Sarno, Hough, and Bulos
(1999) found that 66% of their sample agreed, ‘CCTV makes the public feel safer’,
however, reported feelings of ‘safety’ remained the same whether or not a person was
actually aware of the presence of the cameras (p. 23). In summary, the relationship
between the presence and installation of visible public CCTV systems seems
complex, and somewhat ironically, the presence of CCTV may increase FOC in
some locations.
FOC is a well-researched area by both psychologists and criminologists alike.
Explanations have variously focused on economic vulnerability (Pantazis, 2000),
physical vulnerability (Killias, 1990), media consumption (Gerbner, Gross, Morgan,
& Signorielli, 1994), differential processing of crime messages (Ditton, Chadee,
Farrall, Gilchrist, & Bannister, 2004; Shrum, 2001), social demographic factors
(Farrall, Bannister, Ditton & Gilchrist, 2000), and phobias (Clark, 2004). For a
methodological review of FOC see Ferraro and LaGrange (1987), Farrall, Bannister,
Dittonm and Gilchrist (1997) and Gabriel and Greve (2003); for a literature review
see Hale (1996). In the case of CCTV, the Van der Wurff, Van Staalduinen, and
Stringer (1989) ‘social psychological’ FOC model is relevant because it predicts FOC
by including psychological factors within the individual and aspects of others within
the situation, but also includes aspects of the built environment in which CCTV may
be found.
The Van der Wurff et al. model predicts FOC will increase when potential victims
see themselves or their possessions as an attractive target and they estimate that
others with whom they share public spaces have evil intentions (i.e. they cannot be
trusted because they are likely to be a danger to the self). The model also predicts
that a person’s vulnerability in the face of the power of others affects FOC. Finally,
the model predicts that the location that a person inhabits may seem dangerous
because such things as exit routes may be blocked, or it appears littered with
potential weapons.
According to the Van der Wurff et al. (1989) model, the presence of CCTV within
a location may have quite contradictory effects upon FOC dependent upon whether
a person feels in immediate need of protection. For example, when in no special need
of reassurance, on spotting a camera, the user of a public space might reason that
others have ‘evil intent’ or that the danger presented by powerful others is such that
cameras are called for. Further, CCTV cameras and related signage may suggest an
area especially prone to crime, or frequented by criminals. Mistrust of others in
 Psychology, Crime & Law 745

residential settings, and a low ‘sense of community’ are both associated with FOC
(Sampson & Raudenbush, 1999; Schweitzer, Woo Kim, & Mackin, 1999) and could
be enhanced with the addition of salient deterrent measures in a given location. It
may be that crime deterrence measures like CCTV, as well as offering reassurance,
may act as a situational cue that indicates the extent to which others within a
location pose a danger to the self and ought not be trusted (see Wood, 2006 for a
discussion on this point in terms of surveillance in general). In the long term such
impressions may translate into more enduring estimates of location safety as given in
surveys via some form of affect tagging of recollections of the area (e.g. Slovic,
Finucane, Peters, & MacGregor, 2002, p. 400).
FOC associated with a given location may be a construal governed not just by an
immediate tally of deterrence measures, escape routes and defensible spaces
(Newman & Franck, 1980), but may also include consideration of current goals,
and a complex interpretation of the goals and characteristics, including primed
stereotypical perceptions of others in a given location. The meaning given to a
CCTV camera may differ dependent upon whether it stands over a school
playground (protection for) or a group of youths in a shopping precinct (protection
from). In the former case CCTV may act as a cue that primes notions of vulnerability
and in the latter suggest a threat.
Interpreting the function of the CCTV camera within a given location and the
intentions of those (targets) beneath it may also depend on pre-existing stereotypes
associated with these targets. For example, attributions towards an approaching
person are, of course, likely to be different had this person approached from the
porch of a Church rather than the door of a pub. It is also possible that the gender or
age of the target may interact with a cue like CCTV such that stereotypes such as
‘vulnerable’ versus ‘threatening’ are primed for females and males, respectively, and
that this in turn may affect people’s general impressions of a given situation and
whether it is threatening or not. Finally, despite pre-existing stereotypes and aspects
of the location that may prime concern or offer reassurance, certain targets are in
fact more or less vulnerable within a particular location or appear so dependent
upon their actual behaviour. The lone individual is less of a physical threat to the self
than a large group, and some individuals may appear more intimidating than others
simply because their behaviour tests acceptable conduct in a public space (for
example appearing drunk in public, Dixon, Levine, & McAuley, 2006). For clarity,
the foregoing relationships between stereotypes, targets and a location in terms of
FOC are illustrated in Figure 1.
When out and about people may not always attend to crime deterrence measures
like CCTV in a given location. It is possible that some only attend to such devices
after they have felt threatened by a situation and then find the notion that they may
be being overseen reassuring. The model outlined in Figure 1 assumes that even in
the absence of any ongoing emergency, at least some people notice CCTV (and other
deterrent measures) some of the time. This may not be an unreasonable assumption,
if made with caution. Three months after the installation of cameras in Glasgow city
centre 33% of those sampled within the location knew of their operation (Ditton,
2000). Gill and Spriggs (2005) found levels of awareness to be between 61% and 97%,
although the positive correlation they found between camera density and awareness
was not significant.
746 D. Williams and J. Ahmed

Figure 1. FOC for a specific situation affected by pre-existing stereotypes, target type and the
characteristics of the current location.

In essence, the model predicts a dynamic relationship between salient crime

deterrent measures like CCTV, observer preconceptions and characteristics of a
target to arrive at an overall assessment of a situation as threatening. It is dynamic in
the sense that deterrence cues in a location may prime notions of ‘Evil Intent’ when
there is a correspondence between it and a readily accessible negative social
stereotype for a given target, yet prime positive reassurance in other circumstances.
Context has often been shown to affect people’s interpretation of a range of stimuli,
such as the meaning of letters (Selfridge, 1955), attitudes (Schwarz & Strack, 1991),
and social stereotypes (Bodenhausen, Schwarz, Bless, & Wanke, 1995) including
both gender (Plant, Kling, & Smith, 2004) and race (Graham & Lowery, 2004).
Wittenbrink, Judd, and Park (2001, study 2) measured people’s implicit attitudes
towards either a Black or White target that was shown in either the context of a
‘street corner’ or a ‘church’. A strong interaction was found such that for Black
targets only, negative stereotypic responses were affected by the nature of the context
in the predicted direction. In the present case, CCTV may enhance perceptions of the
‘dangerous’ other if this accords with pre-existing negative beliefs about the threats
which others, especially drawn from certain social groups, may pose. For the
purposes of this study, it is expected that a male ‘skinhead’ may attract more criminal
or antisocial stereotypes, than a ‘studious’ looking female. Stereotypical perceptions
may vary by both gender (Box, 1983; Eagly, 1994; Goodey, 1997; Home Office, 2005,
p. 45) and lifestyle appearance (skinhead/studious), but these are combined here in
order to enhance the differences between the targets in a predictably antisocial/pro-
social direction.
Finally, although the focus has fallen upon CCTV within this introduction, there
are of course many other modern target hardening and situational crime deterrence
measures present in the urban landscape that might also act as stereotype primes or
guide impressions of a location more directly (Clarke, 1995). There is no theoretical
reason why such things as covers to protect other people seeing the entry of pin
numbers for card payments, remote entry systems to flats, swipe card keys, barbed
fencing, toughened barriers in shops, and even the armed officer within an airport,
 Psychology, Crime & Law 747

could not all play the same role. In essence, in respect of FOC for a given location
and trust in the ‘other’, noticeable situational deterrence measures may be a double-
edged sword.

Aims of the study

The present study employs a simple method for assessing the impact of CCTV on
people’s impressions of others and their reported FOC for a given location.
Participants will be shown one background photograph of an urban scene into
which different targets and a CCTV camera can be placed. The placing of a CCTV
camera within the setting is designed to test whether this affects participant ratings of
the location in terms of crime frequency and FOC, but also whether ratings of
targets vary in accordance with stereotypes that might be activated in line with the
function of a CCTV camera. Following this, a male ‘skinhead’, and a ‘studious’
female target will be individually edited into the scene to test whether the former
activates more negative criminal stereotypes in the presence of a CCTV camera than
in its absence, whereas no such difference is predicted for the female ‘studious’ target.
More specifically, it is predicted that a male ‘skinhead’ target shown in the presence
of a CCTV camera will be associated with significantly higher FOC ratings and more
negative impressions than a female target or when either target is shown without a
CCTV camera present.

Method
Participants
There were 120 participants who took part in this study. These were drawn from the
central public shopping area of Hatfield, which is a medium sized (university) town
in Hertfordshire. All participants were recruited between 9.00 am and 5.00 pm over a
2-week period with recruitment sessions taking place both in the morning and
afternoon in equal measure. Potential participants were approached on the basis of
being the next available person passing the researcher once she was ready having
finished with the last person. Participant ages ranged from 18 to 70 years, with the
modal age group being 18
21 years. Fifty-two per cent were female, while 41% were
students, 27% skilled workers, 22% professional employed, 3% homemakers, 3%
unemployed and 4% others. Comparisons that can be made with the 2001 census
data for this area suggests students are overrepresented by 30% in this sample (Office
for National Statistics, 2008). This may reflect the greater willingness of students to
assist in a university-related exercise compared with other groups and that future
work in a similar location might employ quota sampling for this group. All
participants who agreed to take part gave informed consent, and described
themselves as being without any significant visual impairment.

Design
The experiment has a 2 (CCTV: present or absent) 3 (target: type, none, female,
and male) between subjects factorial design. The dependent variables were a picture
location specific crime frequency estimate, a number of ‘walk alone’ FOC estimates
748 D. Williams and J. Ahmed

and a ‘day in the life of’ the target person perception estimate. The experimental
prediction was that a male ‘skinhead’ target shown in the presence of a CCTV
camera will be associated with significantly higher FOC ratings and more negative
impressions than a female target or when either target is shown without a CCTV
camera present.

Materials
In all there were six different urban scenes shown to participants on an A4 sheet of
paper. The pictures were in colour and measured 13 by 22 cm. These pictures appear
in Figure 2. The top row of three pictures shows the scene where CCTV is never
present, while the columns show from left to right, target absent, female target and
male target. In this way CCTV presence/absence is crossed with all target conditions.

Measures
There were four main dependent measures that were directly related to the contents
of the pictures below. Firstly, location-specific crime frequency estimates were made
using the total score of six items adapted from Honess and Charman (1992) which
appeared after the statement: ‘Looking at this particular photograph, how often do
you think each of the following crimes are likely to happen in the next 30 days?’. The
six items that followed were:
1. A violent attack on somebody;
2. An act of vandalism;
3. Drunk and disorderliness;
4. Theft from a person;

Figure 2. The stimulus pictures
top row without CCTV; left column, absent target, middle
column, female target, and the male target is on the right.
 Psychology, Crime & Law 749

5. A sexual assault;
6. A shop break in.

These items were anchored to the following descriptors, ‘very often’, fairly often’,
‘occasionally’, ‘rarely’, ‘not at all’ with a ‘don’t know’ option. The standardized
reliability alpha for these items was 0.73 (M 3.35, SD0.51, high scorehigher
estimated frequency).
Secondly, FOC related to the pictured location was measured using a number of
variations of the ‘walking alone’ question. ‘Looking at this photo’ (very safe, safe,
unsafe, very unsafe, high scoreunsafe), ‘Think about the possible trouble that
could take place, how safe would you feel walking through this street alone at night?’,
‘How about during the day?’, ‘How worried would you be to let a close friend walk
alone through this street?’, ‘How about during the day?’. An overall scale score was
calculated using the total of these questions and dividing by four. This ‘Fear of
walking in the location’ scale had a standardized reliability alpha value of 0.76 (M 
1.91, SD 0.57) with a high score indicating greater fear. Thirdly, a 14-item anxiety
checklist was also used to tap a range of feelings in respect of walking in the location
shown. This additional measure was used in order to avoid mono-operational bias,
a common problem in FOC research (see Ferraro & LaGrange, 1987, p. 76 for a
discussion of common problems encountered with this measure).
The statement, ‘Imagine that you had to walk alone in this area, would this street
scene make you feel?’ was followed by 14 adjectives that could be freely circled. These
were as follows:
Worried, Fearful, Scared, Afraid, Frightened, Threatened, Anxious
Secure, Relaxed, Fearless, At Ease, Reassured, Safe, Comfortable
A total score for this measure was calculated by subtracting the number of secure
items that were checked from the number of fearful items and then adding seven to
maintain a positive scale with a possible range between zero and 14, where higher
values indicate greater fear associated with the location (M3.97, SD 2.85).
Finally, perceptions of the target (where present) were measured by asking
participants: ‘looking at the person in this photograph, could you please write a brief
paragraph of a ‘‘day in the life’’ of this person. Just use your imagination to describe
some of the things this person might do in a day . . . ’. This was followed by eight line
spaces. This method has been used in previous research on stereotype perception
(e.g. Macrae, Bodenhausen, Milne, & Jetten, 1994), where the content of what was
written about the target is coded into preset categories in order to obtain a
quantifiable metric. Two independent raters who were blind to the experimental
conditions coded the passages into the following four categories: anti-social, leisure,
work related or other activities. For example, the statements ‘goes to the gym, looks
active’ and ‘outstays his welcome in the café’ were related to leisure and anti-social,
respectively. A passage could contain statements that were neutral, or countered the
prevailing impression. If such content gave rise to disagreement as to the overall
classification of the passage, this was resolved via discussion.

Manipulation checks
In order to establish whether variations in the presence of a CCTV camera were
correctly reported within conditions, an item was placed at the end of the
750 D. Williams and J. Ahmed

questionnaire asking if this had been noticed (yes, no, don’t know). The results of
this are shown in Table 1. The CCTV camera was noticed in accord with the
experimental manipulation (x2 59.21, d.f.2, p B0.001) Despite this, nine failed to
correctly identify its presence. This may reflect the ease, hindsight apart, with which
even an obvious CCTV camera may blend into a street scene, and so may be difficult
to recall.

Procedure
Potential participants were approached by a female researcher in the main public
shopping area of a medium sized town in Hertfordshire. The main purpose of the
study was described as involving ‘people’s impressions of others and the environ-
ment’. After being given further information, and receiving informed consent,
participants were shown one of the six photos presented in Figure 2 according to the
condition they were in (which ran through the photos in a sequential order of six).
With the photo still in view participants were asked to complete the ‘day in the life
of’ question before they were asked to go onto the crime frequency and FOC
measures which made up most of the rest of the questionnaire. This was in order to
avoid priming participants with crime-related thoughts prior to them giving
impressions of the targets. Where no target was present in the photo, participants
were not given the ‘day in the life’ section prior to completing the questionnaire. In
all cases the photo was removed from sight before participants completed the final
part of the questionnaire that contained the demographic questions and manipula-
tion checks. Finally, all participants were thanked and debriefed at the end of this
procedure.

Results
In order to test the hypothesis a 2 (CCTV present or absent) 3 (target status, none,
female, and male) between-subjects MANOVA using the full sample of 120 (20 per
cell) participants was undertaken. Preliminary analysis found that participant gender
was significantly correlated with one dependent measure (the anxiety checklist, Rho
0.24, n 119, p B0.01) so this variable was entered as a covariate in the MANOVA,
other assumption checks for this analysis were conducted and no serious violations
noted. The multivariate results revealed no main effects for the presence of CCTV
when location crime frequency estimate, fear of walking alone measure, and the
anxiety checklist scores were entered together as dependent measures: F(3,110) 
0.34, NS, Wilks’ lambda 0.99. There was also no significant main effect for the
other independent variable, target status: F(6,220) 0.74, NS, Wilks’ lambda 0.96.

Table 1. The reported detection of CCTV camera in the picture by condition.

Noticed CCTV

CCTV camera Yes No Don’t know Total

Present 39 9 12 60
Absent 2 47 11 60
Totals 41 56 23 120
 Psychology, Crime & Law 751

However, the interaction between both these variables was significant: F(6,220) 
2.23, p B0.05, Wilks’ lambda 0.89, h20.06.
In order to examine the interaction between CCTV and target status in more
detail a further univariate analysis was undertaken. This revealed that of the three
dependent measures entered, only the ‘fear of walking alone’ measure differed
significantly by CCTV and target status: F(2,112) 5.65, p B0.05, h2 0.10 (the
alpha for this test is adjusted using the Bonferroni method). The descriptive statistics
for ‘Fear of walking alone’ composite measure appear in Table 2.
An inspection of the means presented in Table 2 shows that this interaction effect
is such that when CCTV and a male target are present, the ‘fear of walking’ ratings
are highest. A post hoc Scheffé test showed there was no significant difference
between any of the levels of the target status independent variable, however, a t-test
did locate a significant difference (t(38) 2.77, p 0.01) between CCTV present and
males (M 2.16, SD 0.61), versus CCTV absent and males (M 1.66, SD 0.54).
No other comparison was significant.
Manipulation checks reveal that participants did not always correctly report
having seen or not seen the CCTV in accordance with its presence in the photo
(Table 1). Of the 60 participants that were in the CCTV present condition only 39
later correctly report having seen it, with nine saying there were none and 12 who
were unsure. The MANOVA test above was repeated1 using reported awareness of
the CCTV camera (yes, no, or unsure) as an independent variable to replace the
experimental manipulation of its actual presence. The results found no main effect
for CCTV F(4,216) 1.80, NS, Wilks’ lambda 0.94; or target status F(4,216) 
0.46, NS, Wilks’ lambda 0.98; or the interaction between these F(8,216) 0.49, NS,
Wilks’ lambda0.96. This test was also conducted excluding the ‘unsure’ group and
the results were unchanged. This is an interesting finding because it may be that
judgements concerning fear of walking in the location are made online while
participants are still viewing the scene and have no knowledge of which aspects of
the scene they may be asked about later, rather than offline and at the time of
judgement where the effect of contextual factors like salient CCTV would need to be
recalled to influence participants’ impressions of the location or those within it. (e.g.
Bargh, 1997) This follows because the negative impression of the location scene
appears to be formed without later explicit recall of specific components of the scene
that are related to the judgement.
In order to explore whether the impression the target has upon the observer
varies by its apparent threat, a further analysis was undertaken to see if those who
saw the ‘skinhead’ were more likely to recall a CCTV camera when there was none
compared with those who saw either the female or no target1. It is possible that later

Table 2. Fear of walking alone in the location by CCTV and target type.

Target status

CCTV Absent Female Male Total

Absent 2.07 (0.59) 1.89 (0.53) 1.66 (0.54) 1.87 (0.57)

Present 1.71 (0.45) 1.93 (0.54) 2.16 (0.61) 1.93 (0.56)
Total 1.89 (0.55) 1.91 (0.53) 1.91 (0.62) 1.91 (0.57)
Total n119, cell ns 20 or 19, (SD), high scoregreater worry.
752 D. Williams and J. Ahmed

recall is affected by how threatening the target appears within the scene, such that a
CCTV camera is presumed present in these cases when it is in fact absent. The results
of a count by recall error type and target status are shown in Table 3. The frequencies
suggest that there is no association between the presence of a threatening target and
later incorrectly recalling a CCTV camera when one was not present (x2 0.25,
d.f. 2, NS). However, this result is based on a modest sample, which is only present
because of the higher than expected recall error rate. It would be interesting to
investigate this more thoroughly in future work, perhaps by ensuring crime
deterrence measures are always absent, while manipulating target threat using a
range of target features.
Finally, the association between male target gender and the presence of CCTV
was also examined using the target classification categories obtained using the ‘day
in the life’ passage (anti-social, leisure, work related or other activity). The
descriptive results for target classification by CCTV and target status are shown in
Table 4.
A chi-square test of the data in Table 4 revealed that there was a significant
association between CCTV presence and target gender in terms of person perception
(x2 35.19, d.f.9, pB0.001; a separate analysis on both target status and CCTV
presence independent of each other was also carried out, both tables were significant
but are not reported here). Although there is a substantial gender by work-related/
leisure effect, probably due to basic differences in the appearance and demeanour of
the male and female targets, it is still the case that an anti-social classification only
occurs for male targets, and more importantly, only when in the presence of the
CCTV camera.

Summary of results
As expected, when CCTV is placed in the visual context of targets capable of
inducing differing social stereotypes, people’s impressions of the scene and those
within it can change. Self reported concern regarding walking in the photographed
location was highest only when both a male target and CCTV camera were present.
This effect is not associated with later recall of the camera within the street scene, and
target type does not appear to affect later inferences about whether a camera was or
was not likely to be present. Further, descriptions of a ‘day in the life’ of the target
were only anti-social when the target was a male ‘skinhead’ and in the presence of a
CCTV camera. Moreover, no main effects in terms of any dependent measure were

Table 3. CCTV recall errors by CCTV presence and target status.

CCTV condition: CCTV present CCTV absent

Recall: Absent or unsure Present or unsure Totals

Target Status
Male 8 6 14
Female 8 4 12
Absent 5 3 8
Totals 21 13 8
 Psychology, Crime & Law 753
Table 4. Person perception categories by CCTV and target type.

CCTV No CCTV CCTV No CCTV

Categories Male Male Female Female Totals

Anti-social 4 4
Leisure 15 14 7 7 43
Work related 4 13 11 28
Other 1 2 2 5
Totals 20 20 20 20 80

found for target status or CCTV presence, but it is when both these variables
interact, as predicted, that reported FOC and anti-social impressions are highest.

Discussion
As predicted, when a male ‘skinhead’ target is shown in the presence of a CCTV
camera, participants give significantly higher FOC ratings and report more negative
impressions of this person than when a female target is shown or CCTV is absent.
This interaction involves both the presence of a situational crime deterrence device
such as CCTV and the presence of a target more or less likely to prime negative or
an antisocial social stereotypes. The absence of an association between post-
experimental CCTV recall and error type and the FOC measures suggests that the
differential effects found for CCTV are realized via an ‘online’ judgement process
that takes place at the time the scene is viewed. However, the evidence here does not
disentangle a process order. A street context that includes a CCTV camera may
prime accessible antisocial stereotypes, such as an intimidating ‘skinhead’ that would
otherwise not be brought to mind, or an accessible antisocial stereotype may simply
add to the impression that a location is unsafe when viewed with certain other
contextual features. What does seem clear is that there is an interaction between a
salient CCTV camera within a street context and other environmental factors, such
as the presence of others who may or may not easily attract notions of ‘evil intent’
such that an impression of the scene depends upon consideration of both of these
factors together at the time they are seen. In essence, people appear to actively
interpret crime-related cues such as formal surveillance in line with their own pre-
existing expectations of others and the risks they face.
Much of the intuition and discourse that underscores the presence of situational
crime deterrence measures generally, and CCTV specifically, is that they are
inherently reassuring because of their function. Yet for some time now researchers
have begun to question whether this is so. Defending the modern urban landscape
from a sense of undulating moral crisis and corresponding crime with visible
technological crime deterrence measures may not always reduce FOC (e.g. Ditton,
2000; Gill & Spriggs, 2005; Graham, 1998; South, 1987). This study found no main
effect for either CCTV or target type, which calls into question the basic link between
a camera within a street context and reassurance. The interaction found here
suggests that the relationship between these two things is more complex than first
appears. Where pre-existing anti-social stereotypes may be primed, or no pre-existing
sense of threat and immediate need of security are evident, the presence of formal
754 D. Williams and J. Ahmed

deterrence measures like CCTV, when noticed, may in fact come to represent a proxy
symbol of the threat that others pose. Interestingly, Schweitzer et al. (1999) also
found that the density of ‘neighbourhood watch’ signs increased FOC within
American urban locations, so this process may not be specific to CCTV, but part of a
general response to environmental features that can indicate the ‘trustworthiness’ of
others when making FOC-related appraisals.
There is a rich literature linking environmental cues to FOC and trust in others.
For example, the ‘broken window’ effect (Kelling & Coles, 1996; Wilson & Kelling,
1982), environmental physical disorder and ‘incivilities’ (Skogan, 1990; Taylor,
Shumaker, & Gottfredson, 1985), environmental design and layout (Perkins &
Taylor, 1996), the normative behaviour and cohesion of local communities (Sampson
& Raudenbush, 1999), the ‘fortress like’ appearance or isolation of residential
settings (Blakely & Snyder, 1997). What is of interest here is that the environmental
cue in question is partly designed to reduce FOC, and this study demonstrates that in
certain contexts it can have the opposite effect.
Clearly at some point it may be possible to establish an effect in respect of both
CCTV devices and antisocial targets independently of each other given the almost
infinite range of visual images that could be employed to represent each. For
example, one is hardly likely to report a willingness to walk in locations occupied by
a group of armed and extremely threatening others, or where there seems to be
oppressive military surveillance hardware! What is of interest here is the demonstra-
tion of an interaction effect between two visual images insufficiently remarkable to
affect location FOC independently of each other. However, together, the ordinary
urban CCTV camera and male ‘skinhead’ may influence impressions of the overall
scene, and actors within it, in a negative fashion.
It would be reassuring to have concurrent evidence that participants had
attended to those aspects of the image that are implicated. According to the
manipulation checks, most participants correctly recall CCTV presence by condi-
tion, but more detailed evidence of gaze location within the scene would help
support this conclusion, and could be obtained with the use of eye-tracking
methods (Yarbus, 1967). Further, pre-testing participants for target stereotype
content would help strengthen the findings here. It is important to establish that the
two specific target types differ from each other in terms of anti-social stereotype
content, and that this is unrelated to FOC measures before the introduction of
CCTV cues.
Any methodology that employs visual images may be criticized on the basis of
stimuli equivalence. It could be argued that the control condition for CCTV present,
is not its absence, but rather the presence of an equally detailed example of street
technology that does not have the same function as a CCTV camera, such as a
pollution monitor. This could help rule out such things as novelty, interest and
distraction as competing explanations for the effect (Pickel, 1998, 1999) or general
arousal in the presence of a threat cue (Easterbrook, 1959). However, as noted above,
given the almost infinite variety of differences present in visual images, perhaps
again, a better way forward now would be to undertake a conceptually similar test
using video clips, or real settings with a sophisticated means of coding settings for
the presence of cue variables.
Two of the three location FOC measures failed to confirm the predicted
interaction. Recall that a significant interaction was not found using either the
 Psychology, Crime & Law 755

‘location crime frequency’ or the ‘location anxiety checklist’ measures. The former
measure deals with the risk that a particular event may occur in the given location.
This construct is conceptually distinct from personal FOC or victimization (Ferraro
& LaGrange, 1987; Hale, 1996). The latter measure is asking for a current emotional
state related to a picture. With hindsight, it seems unreasonable to expect distinct
current emotional responses to variations in the content of a picture that are
relatively modest. Naturally, ethical constraints limit the extent to which anxiety
states can be experimentally induced, nevertheless, further work might extend the
ecological validity of the methods used here by employing video clips of an area, or
asking participants about their impressions of others or current emotional states in
real situations that differ in terms of crime deterrence cues and user demographics
(a sophisticated example of such ‘systematic observational procedures’ using video
of urban neighbourhoods is provided by Sampson & Raudenbush, 1999). The reason
the ‘worry and safety walking through this street measure’ does respond to pictorial
cues may be because it is a judgement that is highly relevant to the self, as it concerns
a proposed encounter with the situation and those within it. In essence, the
relationship between potential threat cues within the picture begins to matter, when
you or a loved one are going to have to walk though it at some point.
CCTV may fall into a unique category of situational crime deterrence measures
because of its supposed general capacity to reassure through observation, while at
the same time also having the capacity to act as a reminder that others may not be
trusted. Investigating other types of deterrence measures may provide useful and
general insights into the relationship between context, targets, crime cues and FOC.
However, because CCTV is now such a ubiquitous feature of the modern urban
landscape it deserves closer empirical attention on the part of criminal and social
psychologists alike given its potential effect on mutual suspicion, trust and social
cohesion in public spaces. A good deal of research has focused on CCTV
effectiveness in terms of crime reduction and FOC (Ditton, 2000; Gill & Spriggs,
2005; Welsh & Farrington, 2002). The work presented here suggests it may also
negatively interact with people’s impressions of others under certain circumstances,
which in turn could mediate FOC. A recent report for the Information Commis-
sioner, edited by David Wood (2006) describe surveillance as the: ‘purposeful,
routine, systematic and focused attention paid to personal details, for the sake of
control, entitlement, management, influence or protection’ (p. 4) and that for good
or ill this is a current feature of the modern Western state, which includes the general
and routine use of CCTV systems. This report also points out that such surveillance
speaks of ‘a world in which we know we are not really trusted’ (p. 3), yet such trust is
essential for social cohesion and solidarity. Somewhat ironically, these are the very
ingredients often cited as necessary to achieve community-based informal regulation
and order (Sampson & Raudenbush, 1999). The evidence presented here, albeit
confined to a lab-based demonstration, reminds us that modern formal surveillance
may have a range of benefits, but may also have unexpected consequences in terms of
our impressions of others and the spaces they occupy, which in turn may have quite
far-reaching implications for society as a whole.

Acknowledgements
Institutional ethical approval application PSY/12/06/JA.
756 D. Williams and J. Ahmed

Note
1. The authors are grateful to an anonymous reviewer who suggested undertaking this
additional analysis.

References
Bargh, J.A. (1997). The automaticity of everyday life. In R.S. Wyer (Ed.), The automaticity of
everyday life, advances in social cognition (Vol. X) (pp. 1
62). Mahwah, NJ: LEA.
Blakely, E.J., & Snyder, M.G. (1997). Gated communities in the United States. Washington, DC:
Brookings Institute and Lincoln Institute for Land Policy.
Bodenhausen, G.V., Schwarz, N., Bless, H., & Wanke, M. (1995). Effects of atypical exemplars
on racial beliefs: Enlightened racism or generalised appraisals? Journal of Experimental
Social Psychology, 31, 48
63.
Box, S. (1983). Power, crime and mystification. London: Tavistock.
Clark, J. (2004). Crime fear and phobias. Psychiatry. Psychology and Law, 11, 87
95.
Clarke, R.V. (1995). Situational crime prevention. In M. Tonry, & D.P. Farrington (Eds.),
Building a safer society: Strategic approaches to crime prevention: Vol. 19. Crime and justice:
a review of research (pp. 91
150). Chicago, IL: University of Chicago Press.
Ditton, J. (2000). Crime and the city: Public attitudes towards open-street CCTV in Glasgow.
British Journal of Criminology, 40, 692
709.
Ditton, J., Chadee, D., Farrall, S., Gilchrist, E., & Bannister, J. (2004). From imitation to
intimidation: A note on the curious and changing relationship between the media, crime
and fear of crime. British Journal of Criminology, 44, 595
610.
Dixon, J., Levine, M., & McAuley, R. (2006). Locating impropriety: Street drinking, moral
order and the ideological dilemma of public space. Political Psychology, 27, 187
206.
Eagly, A.H. (1994). On comparing women and men. Feminism and Psychology, 4, 513
522.
Easterbrook, J.A. (1959). The effect of emotion on cue utilization and the organization of
behavior. Psychology Review, 66, 183
201.
Farrall, S., Bannister, J., Ditton, J., & Gilchrist, E. (1997). Questioning the measurement of the
‘Fear of Crime’: Findings from a major methodological study. British Journal of
Criminology, 37, 657
678.
Farrall, S., Bannister, J., Ditton, J., & Gilchrist, E. (2000). Social psychology and fear of crime:
Re-examining a speculative model. British Journal of Criminology, 40, 399
413.
Ferraro, K.F., & LaGrange, R. (1987). The measurement of fear of crime. Sociological Inquiry,
57, 70
101.
Gabriel, U., & Greve, W. (2003). The psychology of fear of crime, conceptual and
methodological perspectives. British Journal of Criminology, 43, 600
614.
Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1994). Growing up with television: The
cultivation perspective. In J. Bryant, & D. Zillmann (Eds.), Media effects: Advances in theory
and research (pp. 17
41). Hillsdale, NJ: Erlbaum.
Gill, M., & Spriggs, A. (2005). Assessing the impact of CCTV. Home Office Research Study
292. London: Home Office Research, Development and Statistics Directorate.
Goodey, J. (1997). Boys don’t cry: Masculinities, fear of crime and fearlessness. British Journal
of Criminology, 37, 401
418.
Graham, S. (1998). Towards the fifth utility? On the extension and normalization of public
CCTV. In C. Norris & Armstrong (Ed.), Surveillance, CCTV and social control (pp. 89
112).
Aldershot, UK: Ashgate.
Graham, S., & Lowery, B.S. (2004). Priming unconscious racial stereotypes about adolescent
offenders. Law & Human Behaviour, 28, 483
504.
Hale, C. (1996). Fear of crime: A review of the literature. International Review of Victimology,
4, 79
150.
Home Office. (2005). Home Office statistical bulletin, criminal statistics for England and Wales
19/05. London: RDS Office for Justice Reform.
Honess, T., & Charman, E. (1992). Closed circuit television in public places. Police Research
Group Crime Prevention Unit Series, 35. London: Home Office.
Kelling, G., & Coles, C. (1996). Fixing broken windows: Restoring order and reducing crime in
our communities. New York: Free Press.
 Psychology, Crime & Law 757

Killias, M. (1990). Vulnerability: Towards a better understanding of a key variable in the

genesis of crime. Violence and Victim, 5, 97
108.
Macrae, C.N., Bodenhausen, G.V., Milne, A.B., & Jetten, J. (1994). Out of mind but back
in sight: Stereotypes on the rebound. Journal of Personality and Social Psychology, 67,
808
816.
McCahill, M., & Norris, C. (2003). Estimating the extent, sophistication and legality of CCTV
in London. In M. Gill (Ed.), CCTV. Leicester: Perpetuity Press.
Newman, O., & Franck (1980). Factors influencing crime and instability in urban housing
developments. Washington, DC: Government Printing Office.
Norris, C., McCahill, M., & Wood, D. (2004). The growth of CCTV: A global perspective on
the international diffusion of video surveillance in publicly accessible space. Surveillance &
Society, 2, 110
135.
Office for National Statistics. (2008). Neighbourhood statistics: Welwyn and Hatfield
(Local Authority). Retrieved June 2008 from: http://www.neighbourhood.statistics.gov.uk/
dissemination/LeadDatasetList.do?a3&b276999&cHatfield&d13&g456628&i1001x
1003&m0&r1&s1215172951281&enc1&domainId15
Pantazis, C. (2000). Fear of crime, vulnerability and poverty, evidence from the British Crime
Survey. British Journal of Criminology, 40, 414
436.
Perkins, D., & Taylor, R.B. (1996). Ecological assessments of community disorder: Their
relationship to fear of crime and theoretical implications. American Journal of Community
Psychology, 24, 63
107.
Pickel, K.L. (1998). Unusualness and threat as possible sources of weapon focus. Memory, 6,
277
295.
Pickel, K.L. (1999). The influence of context on the ‘Weapon Focus’ effect. Law and Human
Behavior, 23, 299
311.
Plant, E.A., Kling, K.C., & Smith, G.L. (2004). The influence of gender and social role on the
interpretation of facial expressions. Sex Roles, 51, 187
196.
Sampson, R.J., & Raudenbush, S.W. (1999). Systematic social observation of public spaces: A
new look at disorder in urban neighbourhoods. American Journal of Sociology, 105,
603
651.
Sarno, C., Hough, M., & Bulos, M. (1999). Developing a picture of CCTV in Southwark town
centres. Final Report, Crime Policy Research Unit, South Bank University. Retrieved July
2007 from: http://www.kcl.ac.uk/depsta/law/research/icpr/publications/cctv.shtml#fears
Schweitzer, J.H., Woo Kim, J., & Mackin, J.R. (1999). The impact of the built environment on
crime and fear of crime in urban neighborhoods. Journal of Urban Technology, 6, 59
73.
Schwarz, N., & Strack, F. (1991). Context effects in attitude surveys: Applying cognitive
theory to social research. In W. Stroebe, & M. Hewstone (Eds.), European review of social
psychology (pp. 31
49). London: Wiley.
Selfridge, O.G. (1955). Pattern recognition and modern computers. Proceedings of the 1955
Western Joint Computer Conference: Published by the Institute of Radio Engineers
(pp. 91
93). Los Angeles: Institute of Radio Engineers.
Shrum, L.J. (2001). Processing strategy moderates the cultivation effect. Human Communica-
tion Research, 27, 94
120.
Skogan, W. (1990). Disorder and decline: Crime and the spiral of decay in American cities.
Berkeley, CA: University of California Press.
Slovic, P., Finucane, M., Peters, E., & MacGregor D. (2002). The affect heuristic. In
T. Gilovich, D. Griffin, & D. Kahneman (Eds), Heuristics and biases: The psychology of
intuitive judgment (pp. 397
420). Cambridge: Cambridge University Press.
South, N. (1987). The security and surveillance environment. In J. Lowman, R. Menzies &
T. Pays (Eds), Transcarceration: Essays in the sociology of social control (pp. 129
152).
Aldershot, UK: Gower.
Taylor, R., Shumaker, S., & Gottfredson, S. (1985). Neighborhood-level links between physical
features and local sentiments: Deterioration, fear of crime, and confidence. Journal of
Architectural Planning and Research, 21, 261
275.
Van der Wurff, A., Van Staalduinen, L., & Stringer, P. (1989). Fear of crime in residential
environments: Testing a social psychological model. Journal of Social Psychology, 129,
141
160.
758 D. Williams and J. Ahmed

Welsh, B., & Farrington, D.P. (2002). Crime prevention effects of closed circuit television: A
systematic review. Home Office Research Study 252. London: Home Office Research,
Development and Statistics Directorate.
Wilson, J.Q., & Kelling, G.L. (1982). Broken windows. Atlantic Monthly, 294, 29
38.
Williams, D. (2008). Effective CCTV and the challenge of constructing legitimate suspicion
using remote visual images. Journal of Investigative Psychology and Offender Profiling, 4,
97
107.
Wittenbrink, B., Judd, C.M., & Park, B. (2001). Spontaneous prejudice in context: Variability
in automatically activated attitudes. Journal of Personality and Social Psychology, 81,
815
827.
Wood, D.M. (Ed. 2006). A report on the surveillance society for the information commissioner by
the surveillance society network. London: UK Information Commissioner’s Office.
Yarbus, A.L. (1967). Eye movements and vision. New York: Plenum.